Transmission of CML or of t(9; 22) and BCR/ABL? They are not the same

In a recent Letter to the Editor de Brito et al.1 describe a transplant donor retrospectively found to have t(9;22) and BCR/ABL along with other cytogenetic abnormalities often associated

with acute phase or blastic transformation of CML. There are several interesting points including: (1) the donor had normal blood parameters and no evolution to chronic or acute phase CML

for >6 months (but possibly several years) when t(9;22) and BCR/ABL were detected in a substantial proportion of her myeloid cells; (2) although the donor was chimeric for cells with and

without t(9;22) and BCR/ABL, initial engraftment was with the normal donor cells; and (3) the recipient did not develop chronic or acute phase CML after >6 months follow-up.

Several points warrant consideration. First, it was reported many years ago that persons with chronic phase CML at diagnosis may have CML cells with cytogenetic abnormalities consistent with

acute phase.2 This is conceptually consistent with recent data from ultra-deep sequencing of the BCR/ABL tyrosine kinase domain in persons with chronic phase CML, where many mutations are

identified before drug exposure.3 These data suggest a long latency from acquisition of BCR/ABL to diagnosis of chronic phase CML. Second, having BCR/ABL is not equivalent to having CML.

Data from several centres report normal persons with BCR/ABL.4 The point is that to have CML the BCR/ABL mutation must be in a cell biologically capable of causing the disease. Whether this

was so in this donor (and/or recipient) is unknown. Consequently, starting CML therapy may have been premature and/or unnecessary, especially absent data-treating chronic phase CML earlier

is of benefit. Third, latency from BCR/ABL to developing CML is unknown and likely highly variable. The common notion derived from the A-bomb survivors of 10–15 years latency is probably

wrong. Some children with typical CML, t(9;22) and BCR/ABL, were diagnosed before age 1 year and have a somatic (acquired) mutation implying CML latency can be